Transient behavior of cardiorespiratory interactions towards the onset of epileptic seizures

Abstract

Epileptic seizures are typically related to autonomic dysfunction. During seizures, the cardiac and respiratory mechanisms are deeply affected. This effect of epilepsy can also occur a few seconds before the seizure onset in the EEG. In addition, the interaction between respiration and heart rate is also expected to be affected. This study aims to determine whether the cardiorespiratory interactions change during seizures, and more importantly if they show a transient behavior towards the seizure onset. This is done by means of a time series method based on entropy decomposition applied to ECG and respiratory data. Here, the information carried by the heart rate that can be predicted by its own past, or by the past of the respiration, or by a combination of the two, is quantified. It is shown that cardiorespiratory interactions also change even before the onset of focal, and absence seizures. This suggests that early detection of focal seizures can be improved, and that detection of seizures without a clear effect on the heart rate (i.e. absence) can also be detected. In tonic/tonic-clonic seizures no consistent significant change in the autonomic controls was found.